Data describing the SG management regime were entered into Overseer (Version 6; Web version accessed 13th September 2012), (Table 8.1). A DC system was also modelled in Overseer (hereafter called the MDC1 system). Many of the input parameters for the MDC1 system were based on the field trial e.g. the inclusion of a covered animal shelter as the standoff area, with 8 hours per day grazing, and 100% of the on-farm cows using the animal shelter year-round. In the Overseer simulation, the pad was scraped (without
Chapter 8 – Overseer®
Comparison 149
water) and the effluent kept separate from the farm dairy effluent system. Subsequently, the majority of the ‘MDC1’ farm had animal shelter effluent applied to it, the liquid being sprayed infrequently, and the solids being applied four times per year (November, December, February and April). The quantity of imported supplementary feed and manufactured fertiliser, cow numbers and milksolids production were the same as values used in the SG system (Table 8.1).
In a second exercise, the inputs to Overseer were modified so as to more closely represent effluent return to the DC plots in the field trial (MDC2). This allowed for an assessment of the agreement between an Overseer prediction of N losses from a DC grazing system and measured values.
Table 8.1: Parameters entered into Overseer (based on Massey No. 4 Dairy Farm) for both SG and
MDC treatment systems
Cow numbers / Breed / Total farm area 630 / Friesian x Jersey / 211 ha
Total milk solids production 226800 kg/yr
Supplements Imported (all on DM basis)
(SG fed on feed pad for 1 hour per day
during lactation;
DC fed in covered animal shelter with 8
hours grazing per day)
Maize silage (180 t/yr) Triticale silage (90 t/yr) Pasture silage (120 t/yr) Hay (15 t/yr)
Palm kernel meal (40 t/yr)
Average rainfall 980 mm/yr
Soil type / drainage Tokomaru silt loam / mole & pipe drained
Fertiliser inputs (kg/ha; month)
Sulphur super 30 (200 kg/ha; March) Urea (38 kg/ha; April)
Ammonium sulphate (71 kg/ha; October) Ammonium sulphate (71 kg/ha; November) Urea (76 kg/ha; December)
Winter management SG and DC 40% cows off 1st June to 31st July
Soil tests
(Assuming the same values over whole farm, from soil tests performed on trial plots)
Olsen P (mg/kg) QT K SO4-S (mg/kg)
150 Chapter 8 – Overseer® Comparison
8.4 Results and Discussion
Measured TN leaching losses from the SG treatment were 18, 13 and 26 kg N/ha/yr (average of 19 kg N/ha) for the years 2009, 2010 and 2011, respectively. The variation in leaching values across the 3 years is partly explained by differences in drainage patterns. Although total drainage amounts were similar over the 3 years (373, 316 and 329 mm), the initiation and length of the drainage seasons varied. Furthermore, the concentrations of NO3--N in drainage water, making up on average 64% of the TN concentration, varied
markedly with length of drainage season (Chapter 4). This was expected due to the seasonal nature of N availability and pasture uptake, and the coincidence of grazing and drainage events (Houlbrooke 2005).
Total N leaching losses of 13, 7 and 14 kg N/ha/yr (average 11 kg N/ha/yr) were measured from the DC treatment. Compared to the SG treatment, this translates into a reduction in TN leaching of 28%, 46% and 46% over the three years, 2009, 2010 and 2011, respectively. This equated to a 42% reduction on average, which exceeds the Dairy Industry’s goal of a 30% reduction in N leaching losses (DairyNZ 2010).
When the surface runoff losses (3.2 kg N/ha for SG and 2.9 kg N/ha for DC) were added to the leaching losses, the average overall TN losses to water for the SG treatment was 22 kg N/ha/yr, compared with 14 kg N/ha/yr for the DC treatment (a 36% reduction). Therefore, on average 83% of TN losses were through drainage, which is consistent with the values measured by Houlbrooke (2005) at the same site.
When the SG treatment data were entered into Overseer, the N loss to water for SG grazing was estimated to be 23 kg N/ha/yr, which was slightly higher than the 3-year measured average value for the SG plots (22 kg N/ha/yr). While these losses were very close, Overseer predicted that all of the N lost to water was as leaching and no N was lost as runoff, whereas 3.2 kg N/ha was measured in runoff in the field study.
In order to predict N losses to water, Overseer simulates N transfer via dung, urine and effluent streams and predicts the N loss outcome associated with grazing and effluent management. The DC grazing strategy reduces the urine deposition to paddocks and uniformly applies the captured excreta as a slurry or effluent – all of these processes can be simulated within Overseer’s model framework.
The Overseer simulation of the MDC1 system predicted an average TN loss to water of 14 kg N/ha/yr, which was, in general terms, similar to the average loss measured in the trial
Chapter 8 – Overseer®
Comparison 151
for this treatment. Again, Overseer predicted that all of the loss to water was as leaching, with no runoff losses of N, compared to a 2.9 kg N/ha runoff loss measured in the study.
In the Overseer prediction for the MDC1 system, the annual average effluent application was 176 kg N/ha/yr. It is notable that the theoretical (Salazar et al. 2010) and measured effluent return to the DC plots averaged 136 and 109 kg N/ha/yr, respectively (Table 8.2). However, the long-term average return of effluent N predicted by Overseer at 176 kg N/ha/yr was 40 and 67 kg N/yr above the theoretical target and measured slurry returns.
Table 8.2: The estimated annual return of nitrogen in dung and urine to DC and SG treatments, the
theoretical target slurry application, the difference in N returned to DC plots, the actual return of N in slurry, and Overseer predictions (kg N/ha/yr).
Year Amount N estimated returned as dung and urine1 Theoretical target slurry application1,2 Fertiliser applied (3yr annual average) Theoretical Total N returned Theoretical difference in N returned to DC plots Actual N application in form of slurry 2008/09 SG 383 60 443 DC 155 136 60 351 -92 2123 2009/10 SG 340 100 440 DC 137 121 100 358 -82 nil 2010/11 SG 425 85 510 DC 172 151 85 408 -102 115 Ave SG 383 82 464 Ave DC 155 136 82 372 -92 109 Overseer estimate 176 82
1Estimated using modified version of model by Salazar et al. (2010) based on number of grazings in each
year and average herbage N concentrations.
2Assuming 45% losses of N from animal shelter, based on Longhurst et al. (2006). 3
Biennial strategy of slurry application.
One possible reason for the difference between estimated and measured N returns in effluent/slurry is that Overseer may assume smaller N losses in storage than the Longhurst et al. (2006) values, which were used in the theoretical calculations. Nitrogen losses during storage are difficult to predict, not least of all because of variation in storage time between farms.
152 Chapter 8 – Overseer® Comparison
In the second exercise, the Overseer simulation of the MDC system was modified (MDC2) to deliver 109 kg N/ha/yr as effluent (i.e. to approximate the DC treatment at the field trial). This was achieved by keeping all other parameters the same as the MDC1 system but increasing the covered storage time of the animal shelter effluent solids to 36 months. With an effluent application of 109 kg N/ha, Overseer predicted TN losses to water from the system at 10 kg N/ha/yr from leaching and 0 kg N/ha/yr from runoff, compared with the field measurements of 11 kg N/ha/yr from leaching and 3 kg N/ha/yr from runoff. This result shows that when the N returned to pasture in effluent was adjusted in Overseer to equal the average returns in the trial (109 kg N/ha/yr), then the N leaching predicted by Overseer was in good agreement with the measured values. However, Overseer did not predict the runoff N losses measured in the trial. Runoff is inherently difficult to model, particularly from annual average inputs, due to the different factors that influence these losses.
There is general agreement between N losses to water in Overseer and the results seen in the field trial. Overseer was able to predict the N leaching loss reductions achievable using the DC mitigation strategy, which demonstrates that it has been partially updated to reflect the results from research of mitigation strategies (Overseer 2012), such as using animal shelters. An important point to take from this work is that measured N loss to water varied by ca. 50% for both treatments between the three years. Overseer’s attempt at predicting the long-term average N loss to water for the SG treatment was within 1 kg N/ha/yr, and the model did predict the large reduction from DC plots, within 4 kg N/ha/yr. While 4 kg is a large percentage difference in this situation, the important aspect remains the large reduction caused by the mitigation, and reflected in the Overseer predictions.
However, there remains uncertainty surrounding the simulation of N returned as stored effluent and N losses in runoff. It is important for the quantities of effluent N returned to pastures in DC grazing systems to be accurately estimated because a much larger proportion of nutrients are returned to pastures as effluent compared to standard grazing practice. Further, farm-scale research of DC grazing systems is required to provide improved understanding of nutrient dynamics in stored effluent and the quantity of nutrients actually returned to pastures in these systems.
Chapter 8 – Overseer®
Comparison 153
8.5 Conclusion
In a 3 year field trial at Massey University, DC grazing resulted in a 36% reduction in TN losses (leaching plus runoff) relative to a standard dairy grazing system. When a duration- controlled system was modelled using Overseer (MDC1), the predicted reduction in TN losses was 39% compared with a standard grazing system. When DC grazing as practised at the field trial was simulated with Overseer (MDC2), there was good agreement with measured values for N leaching.
Overseer (Version 6) seems to predict N leaching losses under DC grazing reasonably well. Issues that require further consideration in Overseer relate to the proportion of TN losses in runoff, and the quantity of N returned to pastures in effluent collected from stand- off facilities.
8.6 References
Blakemore, L.C.; Searle, P.L.; Daly, B.K. 1987. Methods for chemical analysis of soils. New Zealand Soil Bureau Scientific Report 80.
DairyNZ 2010. Dairy Industry Strategy - directions to the vision. Hamilton, NZ.
Hewitt, A.E. (Ed.) 1998. NZ Soil Classification (2 edn). Manaaki Whenua - Landcare Research New Zealand Ltd Press: Lincoln.
Hosomi, M.; Sudo, R. 1986. Simultaneous determination of total nitrogen and total phosphorus in freshwater samples using persulfate digestion. International Journal of Environmental Studies 27(3/4); 267-275.
Houlbrooke, D.J., 2005. A study of the quality of artificial drainage under intensive dairy farming and the improved management of farm dairy effluent using 'deferred irrigation'. PhD Thesis. Massey University, Palmerston North.
Houlbrooke, D.J.; Horne, D.J.; Hedley, M.J.; Hanly, J.A.; Scotter, D.R.; Snow, V.O. 2004. Minimising surface water pollution resulting from farm-dairy effluent application to mole-pipe drained soils. I. An evaluation of the deferred irrigation system for sustainable land treatment in the Manawatu. New Zealand Journal of Agricultural Research 47; 405-415.
Longhurst, R.D.; Luo, J.; O'Connor, M.B.; Pow, T. 2006. Herd Homes: nutrient management and farmer perceptions of performance. Proceedings of the New Zealand Grassland Association 68; 309-313.
Monaghan, R.M.; de Klein, C.A.M.; Muirhead, R.W. 2008. Prioritisation of farm scale remediation efforts for reducing losses of nutrients and faecal indicator organisms to waterways: A case study of New Zealand dairy farming. Journal of Environmental Management 87; 609-622.
154 Chapter 8 – Overseer® Comparison
Overseer (2012). Changes to the pastoral N model - Technical Note No. 5. Overseer (2013). www.overseer.org.nz, accessed 26th November 2013.
Parliamentary Commissioner for the Environment 2012. Water Quality in New Zealand: Understanding the Science. Parliamentary Commissioner for the Environment, Wellington, NZ.
Salazar, M.E.; Hedley, M.J.; Horne, D.J. 2010. Using turnips to reduce soil K loading on the effluent block. Proceedings of the New Zealand Grassland Association 72; 247- 250.
Chapter 9 – Summary 155